Messages in this thread | ![/](/images/icornerl.gif) | | Subject | Re: As 2.0 looms | Date | Sun, 28 Apr 1996 17:24:26 +0200 | From | Alain Knaff <> |
| |
>Alain KNAFF: >> >You decide >> >which one you want to use at link time (default is to ignore LD_xxx - >> >it's not necessary except for testing new shared libraries). I think >> >> LD_xxx is also needed for programs such as zlibc, which work by >> redefining certain library calls. Ignoring LD_xxx by default would >> make any such program impractical. > >OK, thanks for pointing this out. But you can still make the binaries >use LD_xxx if you want.
Yes, but my point is that ignoring LD_xxx shouldn't be the default. The main "selling point" of zlibc is that it is usable without recompiling any binaries. This would of course fail if the default would be to ignore LD_xxx (and thus any binaries from distributions would ignore it)
> >There should be a secure LD_PRELOAD equivalent - use a system-wide config >file instead of environment variables. I think David Engel is working on >it (specifying libraries to preload in /etc/ld.so.preload). Is it OK I know :) It was me actually who sent him that patch :)
>for zlibc, or does every user really have to be able to redefine library >calls on their own? (/etc/ld.so.preload would be useful for other things >as well, like replacing crypt() with something better.) The system wide /etc/ld.so.preload is ok if zlibc is installed system wide. However, LD_PRELOAD allows users without root access to install zlibc on their own in their home directory.
> >Marek >
Alain
-----------------------------------------------------------------------. Email: Alain.Knaff@inrialpes.fr Alain Lucien Knaff . Tel.(work): (33) 76 61 52 68 . Tel.(home): (33) 76 85 23 05 Appartement 310b . (répondeur & minicom 3612) =====O=====/ 11,rue Général Mangin. Fax : (33) 76 61 52 52 =====O=====/ 38100 Grenoble France.
| ![\](/images/icornerr.gif) |